Double resnet50 batch size in benchmark script (#17416)
authorJunjie Bai <bai@in.tum.de>
Fri, 22 Feb 2019 23:01:46 +0000 (15:01 -0800)
committerFacebook Github Bot <facebook-github-bot@users.noreply.github.com>
Fri, 22 Feb 2019 23:30:35 +0000 (15:30 -0800)
commit807632d40275586f2cddef8f7314b2011f288bfc
treef399318b1028f9446d863bd1f3a8183305025ee9
parent6d744f8fbfe6ef1c90534893d4ca2b2caf8a0df4
Double resnet50 batch size in benchmark script (#17416)

Summary:
The benchmarks are now running on gpu cards with more memory
Pull Request resolved: https://github.com/pytorch/pytorch/pull/17416

Differential Revision: D14190493

Pulled By: bddppq

fbshipit-source-id: 66db1ca1fa693d24c24b9bc0185a6dd8a3337103
.jenkins/caffe2/bench.sh